English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We tackle the challenging issue of aggressive fine-tuning encountered duringthe process of transfer learning of pre-trained language models (PLMs) withlimited labeled downstream data. This problem primarily results in a decline inperformance on the subsequent task. Inspired by the adaptive boosting method intraditional machine learning, we present an effective dynamic correctiveself-distillation (DCS) approach to improve the fine-tuning of the PLMs. Ourtechnique involves performing a self-distillation mechanism where, at eachiteration, the student model actively adapts and corrects itself by dynamicallyadjusting the weights assigned to individual data points. This iterativeself-correcting process significantly enhances the overall fine-tuningcapability of PLMs, leading to improved performance and robustness. Weconducted comprehensive evaluations using the GLUE benchmark demonstrating theefficacy of our method in enhancing the fine-tuning process for various PLMsacross diverse downstream tasks.

Dynamic Corrective Self-Distillation for Better Fine-Tuning of Pretrained Models